Various methods are known for combining signals from different types of sensors in order to distinguish fires from nuisance conditions. These multicriteria detectors are intended to sense the earliest fire products in order to achieve a quick and accurate response. Representative forms of processing are disclosed in each of Lee D. Tice, Fire Detection System and Method Using Multiple Sensors, US 2006/0119477 A1, Jun. 8, 2006; Lee D. Tice, Multi-Sensor Device and Methods for Fire Detection, US 2006/0181407 A1, Aug. 17, 2006; Lee D. Tice, Multi-Sensor Device and Methods for Fire Detection, US 2004/0189461 A1, Sep. 30, 2004; Lee D. Tice, Multiple Sensor Apparatus and Method, U.S. Pat. No. 5,483,222, Jan. 9, 1996; Lee D. Tice, Apparatus Including a Fire Sensor and a Non-Fire Sensor, U.S. Pat. No. 5,659,292, Aug. 19, 1997; Lee D. Tice, Multi-Sensor Device and Methods for Fire Detection, U.S. Pat. No. 7,068,177 B2, Jun. 27, 2006, assigned to the assignee hereof and incorporated by reference.
Currently, in known systems, end-users can only select a detector sensitivity (or in rare cases specify descriptive labels, such as “lobby”). End-users are often not well equipped to properly select a sensitivity for a given space.
In addition, performance based codes are becoming more prevalent. Performance based codes provide placement of detectors based on system performance goals (such as building occupants evacuated in 60 seconds) rather than prescriptive requirements (such as one detector every 30 feet). As the change is made to performance based codes, fire detection systems that support detectors that are aware of their environment will become more valuable.
It has also been recognized that when smoke detectors are positioned in a high airflow area (such as near a vent), their response can be delayed or entirely disabled. Furthermore, other ambient conditions can make false alarms more likely (c.f. shower steam).
There is thus an on-going need to be able to incorporate environmental information into alarm decision processing so as to enhance performance and to improve reliability. It would be preferable if such information could be acquired and updated, automatically, to the greatest extent possible while the system is carrying out normal processing.